0436: Diagnostic value of serial ultrasensible troponin I measurements for acute coronary syndrome identification in cardiac intensive care unit

Introduction ACS diagnosis relies on the detection of an increase or a decrease in troponin levels (delta-TnI) that might be estimated as absolute (delta-TnI-abs) or relative value (delta-TnI-rel). The discriminative cut-off value of delta-TnI to identify ACS in patients admitted in ICU is a matter of debate. Methods We aimed to determine in a monocentric registry the cut-off value of delta-TnI to identify ACS among patients admitted in ICU with a suspicion of acute myocardial ischemia. All the patients had 2 measurements of ultrasensible cTnI (ADVIA Centaur Siemens®; 99th percentile: 0,04 μg/L). Patients were classified in 4 groups: 1) ACS (NSTEMI, UA), 2) Cardiac non coronary diseases (heart failure, PE, arrhythmia, pericarditis, myocarditis, syncope, Tako-Tsubo), 3) Non cardiac diseases (pneumonia, stroke, epilepsy, neoplasm), 4) Chest pain of unknown origin (musculo-skelettal pain…) Results 275 patients were included. Medians of delta-TnI-abs were 0.12 [0 ; 1.01] ; 0.01 [0 ; 0.03] ; 0 [-0.01 ; 0] et 0 [0 ; 0] μg/L in groups 1 to 4 respectively. Medians of delta-TnI-rel) were 104 [0 ; 680] ; 2 [0 ; 62] ; 0 [-6 ; 0] et 0 [0 ; 0] % in groups 1 to 4 respectively. The delta-TnI-abs was significantly different between groups 1 and 2 (p=0.026). ROC curve analysis showed that optimal threshold of delta-TnI-abs was 0.02 μg/L (sensitivity 67%, specificity 76%) to distinguish group 1 from groups 2, 3 and 4. The threshold was 0.04 μg/L (sensitivity 61%, specificity 80%) to distinguish group 1 from group 2. The use of delta-TnI-rel yielded similar results and the only confounder was age. Conclusion This study suggests that the cut-off of delta-TnI to identify ACS in ICU patients should be higher to be discriminative of cardiac non coronary diseases. The identification of type 1 and type 2 myocardial infarction is a major challenge and needs further research.